[cd0426]

Hello I need an example line of code to do a G03.

I have a Fanuc O-M controller.

I want to make a 1.25" circle with the with a .75" endmill.

Can not figure out how to calculate the I & J value or the order to put them in.

Thanks,
CD

[beege]

I and J are the distance in X and Y that the center point exists relative to the start point. For example, if you have a point at X.5 Y2. where a circle starts and the circle center is at X.75 Y2., your code would look like:

G01 X.5 Y2.
G03 X.5 Y2. I.25 J0

because the center point is .25 from the start point in X, and no distance from the start point in Y. Hope this is understandable.

I'm not giving you the answer for two reasons. First, the teach a man to fish thing, and second, I don't know where your circle exists on your coordinate system. Good Luck!

[guru]

I and J are the distance to center of the Arc in X and Y. Fanuc will read I and J as incremental distance.

So if you start your arc at 3 O'clock... "I" would be minus "J" would be 0(zero)

If you start your arc at 11 O'clock.."I" would be positive and "J" would be negative

[guru]

Quick sample codes for you...

starting at 9 O'clock

(G41)G1X-1.0 Y0 F... (feed in to edge of circle)
G3 X-1.0 Y0 I1.0 J0 ( do full circle, X and Y end points can be left out)
(G40)G1 X-1.5 (move away)

[BlueChip]

Try this link :
http://www.kentechinc.com/tip7.html

Look for the free circle milling code generator on that page ... it will come in handy for ya.

Real World Machine Shop Software at ... www.KentechInc.com

[CNCRim]

Let me jump in.
Is it must I J? If not you can alway just used simple R instead it's just work as good.

[fanuc-support.c]

CNCRIM is right just use R

I=the distance from the centre point of circle to the start in X with direction
J=the distance from the centre point of circle to the start in X with direction

Simple.

PM me and i'll send you a sheet explaining it all.

[sinha_nsit]

But, cannot make a complete circle with R method. The reason is that, for the same start and end points, infinite circles are possible, all with the same radius. In such a case, the circle would need to be made at least in two halves, if R method is desired to be used.
While both methods can be used, selection should generally be based on how dimensions are specified on the drawing.

[CNCRim]

But, cannot make a complete circle with R method. The reason is that, for the same start and end points, infinite circles are possible, all with the same radius. In such a case, the circle would need to be made at least in two halves, if R method is desired to be used.
While both methods can be used, selection should generally be based on how dimensions are specified on the drawing.

what are you talking about? if you want to make an arc more 180deg all you need is put a munis sign in from of the number like R-x.xxxx.
Will make complete 1" circle.
G1X0Y.5
G3Y.5R-.5

[sinha_nsit]

See the attached figure.
For the coincident start/end points, infinite circles are possible, all with the same radius. Possibly the machine would make one, which may or may not suit you.

[fanuc-support.c]

But, cannot make a complete circle with R method. The reason is that, for the same start and end points, infinite circles are possible, all with the same radius. In such a case, the circle would need to be made at least in two halves, if R method is desired to be used.
While both methods can be used, selection should generally be based on how dimensions are specified on the drawing.

WRONG (nuts)

Again i agree with CNCRim

[sinha_nsit]

Have you looked at the figure posted by me?
It is not that a complete circle cannot be made. But the question is: which circle (see the figure)?

[CNCRim]

Have you looked at the figure posted by me?
It is not that a complete circle cannot be made. But the question is: which circle (see the figure)?

Good point, but we all know that is easiest point of a circle is at 0 90 270 and 360 deg relate to the circle center ....... We just trying to show the fast way to program efficently and it is R, calculated I J is take longer except full circle. Additional note, if your control don't like complete full circle with R like the last post I had and you can help it by program it slicely off by .00005. Can you measure or see the circle is out of round by fifty thousand? I don't think so.

G1X0Y.5
G3x.00005Y.5R-.5

[sinha_nsit]

G1X0Y.5
G3x.00005Y.5R-.5

Good idea.
However, X0.00005 and X-0.00005 would make different circles.
The desired circle can be made by manipulating the signs of X and/or Y in G02/G03 block.

[CNCRim]

However, X0.00005 and X-0.00005 would make different circles.
/G03 block.

That's programmer job to know which circle he/she want, you don't think a company pay you for nothing, addition that's what make difference you are experience programmer.

[SBC Cycle]

what are you talking about? if you want to make an arc more 180deg all you need is put a munis sign in from of the number like R-x.xxxx.
Will make complete 1" circle.
G1X0Y.5
G3Y.5R-.5

Just FYI, not all controllers will evaluate a full arc with the R designation. My Fadal requires I and J when interpolating a full arc. There are very valid reasons for this, which sinha_nsit has illustrated. I couldn't locate a Fanuc manual online but can you verify that the Fanuc control will evaluate a full arc using an R address? If it can, how does the control evaluate where the arc center is??

That's programmer job to know which circle he/she want, you don't think a company pay you for nothing, addition that's what make difference you are experience programmer

Exactly. I would think a company would pay an experienced programmer to be able to calculate an arc center using IJ. (nuts)

R values are a shortcut. The control looks at the arc radius value, the arc direction, and the arc sweep (degrees) to calculate the I and J for you. Simply put, it can't do that without additional information, such as the direction of the cut before it encounters the arc move. They definitely have their place for expediency (lathe programming comes to mind) but they aren't a substitute for robust programming. Teaching a guy that doesn't grasp the IJ concept to use R is a disaster waiting to happen.

[SBC Cycle]

CNCRIM is right just use R

I=the distance from the centre point of circle to the start in X with direction
J=the distance from the centre point of circle to the start in X with direction

Simple.

PM me and i'll send you a sheet explaining it all.

Not even close. The J for X value was probably just a typo but there is a major difference between the vector quantity from the center of a circle to the start, as opposed to the vector quantity from the start to the center. The magnitude (distance) is the same but the direction is not.

[sinha_nsit]

Yes. Center method is the basic method. R method has been made available for our convenience. R method, however, cannot be used in certain cases where it may not work or may give incorrect/inaccurate results. For example, as per 0i manual, it should not be used when the arc is close to 180 degree since "the calculated center coordinates may contain an error. In such cases, specify the center of the arc with I, J and K".